seedance-prompt

To write effective AI video generation prompts, use specific physics language and measurable degree adverbs instead of vague mood words. The Director's Formula structures your prompt with concrete details for Subject, Action, Scene, Camera, Style, and Constraints.

The Director's Formula for text-to-video prompts structures instructions into six elements: Subject, Action, Scene, Camera, Style, and Constraints. This structure ensures the AI video generator receives concrete, production-ready guidance rather than ambiguous instructions.

Creating image-to-video prompts requires different strategies than text-to-video scenarios. A dedicated gate differentiates the two, adjusting the prompt structure and constraints to suit the specific input scenario and ensure the AI video generator interprets the source material correctly.

Your AI video generation prompt should be between 30 and 100 words. This target length is enforced to optimize prompts for Seedance 2.0, ensuring the instructions are concise enough for the AI to process while providing sufficient detail for generation.

To adapt your prompt structure for different AI video categories, genre routing adjusts the format based on the content type. It tailors the prompt for categories like Product, Lifestyle, or Drama, ensuring the visual language matches the desired intensity and context.